SciELO - Scientific Electronic Library Online

vol.110 issue5-6A review on opportunities for the development of heat pump drying systems in South AfricaClimate and the mfecane author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand



Related links

  • On index processCited by Google
  • On index processSimilars in Google


South African Journal of Science

On-line version ISSN 1996-7489
Print version ISSN 0038-2353

S. Afr. j. sci. vol.110 n.5-6 Pretoria Mar. 2014




Review of carbon dioxide capture and storage with relevance to the South African power sector



Khalid OsmanI; Christophe CoqueletII; Deresh RamjugernathI

ISchool of Chemical Engineering, University of KwaZulu-Natal, Durban, South Africa
IIMINES ParisTech, Centre Thermodynamic of Processes (CTP), Fontainebleau, France





Carbon dioxide (CO2) emissions and their association with climate change are currently a major discussion point in government and amongst the public at large in South Africa, especially because of the country's heavy reliance on fossil fuels for electricity production. Here we review the current situation regarding CO2 emissions in the South African power generation sector, and potential process engineering solutions to reduce these emissions. Estimates of CO2 emissions are presented, with the main sources of emissions identified and benchmarked to other countries. A promising mid-term solution for mitigation of high CO2 emissions, known as CO2 capture and storage, is reviewed. The various aspects of CO2 capture and storage technology and techniques for CO2 capture from pulverised coal power plants are discussed; these techniques include processes such as gas absorption, hydrate formation, cryogenic separation, membrane usage, sorbent usage, enzyme-based systems and metal organic frameworks. The latest power plant designs which optimise CO2 capture are also discussed and include integrated gasification combined cycle, oxy-fuel combustion, integrated gasification steam cycle and chemical looping combustion. Each CO2 capture technique and plant modification is presented in terms of the conceptual idea, the advantages and disadvantages, and the extent of development and applicability in a South African context. Lastly, CO2 transportation, storage, and potential uses are also presented. The main conclusions of this review are that gas absorption using solvents is currently most applicable for CO2 capture and that enhanced coal bed methane recovery could provide the best disposal route for CO2 emissions mitigation in South Africa.

Keywords: carbon dioxide; capture; storage; emissions; South Africa




There has been a nearly 100% increase in worldwide CO2 emissions since 1971. This increase is of great concern to scientists, governments and the public in general as there is general consensus from the greater scientific community that CO2 - a greenhouse gas - is one of the main contributors to rapid climate change1 experienced globally, especially in the last few decades.

Globally, 78-83% of CO2 emissions can be attributed to electricity generation from fossil fuels.2 In South Africa the situation is no different with almost 93% of the country's electricity needs provided by fossil fuels; 77% of electricity is provided specifically by coal power plants.3,4 Because of the country's abundant coal reserves, the use of relatively inexpensive coal-derived power is unlikely to cease in the next 200 years.3 Coal power plant operations have resulted in South Africa's power sector being the 9th highest CO2 emitting power sector in the world, with an estimated 218 mega tonnes (Mt) of CO2 emitted each year.4,5 Although these emissions are low compared to more developed countries, they are higher than the next nine countries in Africa combined.4 Eskom Ltd., the country's primary electricity utility, is currently the 2nd highest CO2 emitting company in the world, as a result of its utilisation of pulverised coal (PC) combustion plants.

As can be seen in Figure 1, the most significant CO2 emission sources in South Africa are situated in the Gauteng, Mpumalanga and Free State Provinces - not surprisingly, as these provinces form the heart of South Africa's coal mining sector and the regions in which most coal power plants are situated. Figure 1 shows not only the CO2 emissions from coal power plants, but also those from coal-to-liquids industries, gas-to-liquids industries and oil refining processes.



In an effort to reduce CO2 emissions and encourage a move towards a cleaner energy strategy, the South African government is considering proposing a CO2 emissions tax that would be levied on all CO2 emission sources. Recent debates have suggested a tax rate of R75 to R200 per tonne of CO2 emitted; with the most recent and currently applicable cost being R120/tonne CO2 in line with international standards.7 Considering that South Africa's energy industry emits well over 200 Mt of CO2 per annum, the proposed levy will result in significant increases in operating costs for companies in this sector. With Eskom having over the last few years almost doubled its electricity tariff, the proposed CO2 emission tax will further add to the need for Eskom to increase its tariff if it passes on this cost to the consumer. It is therefore imperative that solutions to reduce CO2 emissions be found.

Carbon capture and storage (CCS) is a promising mid- to long-term solution to reduce CO2 emissions. This strategy involves capturing CO2 at power plants and other industries before they are emitted, transporting CO2 to suitable disposal locations, and either storing CO2 underground or utilising CO2 to retrieve high-value products.

In this review, we concentrate on coal power plant operations and their suitability for CCS technology. Techniques that are potentially applicable to CO2 capture in coal power plants are presented, and CO2 transportation, storage and potential uses are discussed with specific relevance to South Africa.


Coal power plant operations

Currently, South Africa possesses 14 PC power plants; 7 of them are in the top 30 highest CO2 emitting power plants in the world.3,5

A simplified schematic of a typical PC power plant is shown in Figure 2. In a PC power plant, coal is transported to a pulveriser via conveyor belts and crushed into a powder with a particle diameter of approximately 50 µm. Hot air then blasts the coal into a boiler where it is burnt. The heat generated is used to heat tubes containing water. These tubes can be kilometres long, but are coiled in order to be compact.8 The water in the heat exchanger tubes is converted into superheated steam at high pressure. The steam is used to drive turbine blades which spin the turbine. The turbine shaft is linked to a generator rotor, which generates electricity using an electromagnet.8 The electricity flows through transmission lines and transformers to reach consumers at the required voltages. The used steam is then cooled and condensed in cooling towers, and recycled to the boilers for reheating.



The gases that are released during the coal combustion are filtered using bag filters to remove ash. If the gas mixture contains substantial sulphurous and nitrogenous emissions, particularly SOX and NOX compounds, then desulphurisation and denitrification processes can be installed to remove them, although these processes have currently not been implemented in South African coal power plants. The remaining gases are emitted through a stack as flue gas. Flue gas composition varies according to coal composition and power plant flue gas treatment processes. The typical composition of flue gases is approximately 12-12.8% CO2, 6.2% H2O, 4.4% O2, 50 ppm CO, 420 ppm NOx, 420 ppm SO2 and 76-77% N2. The flue gas is typically emitted at pressures ranging from 100 kPa to 170 kPa and temperatures of 363.15-412.15 K.9-11

PC power plants typically emit CO2 on a magnitude of 514 kg CO2/MWh electricity produced.12

CO2 removal from PC power plants entails retrofitting the power plant with a CO2 capture process to treat the flue gas for selective CO2 removal before it is emitted through the stack. This mode of CO2 capture is known as post-combustion capture, because CO2 capture occurs after coal combustion.

PC combustion is a well-developed and common power plant process that requires a lower investment cost compared to newer technologies. However, CO2 capture and compression is expensive as the flue gas to be treated is available at unfavourably low pressure and high temperature.


Techniques of capturing CO2 from pulverised coal power plants

Currently, there are many gas separation techniques under investigation for post-combustion CO2 capture from PC power plants. This section explains the unique properties of CO2 and presents CO2 capture techniques which exploit these properties for efficient gas separation, despite the unfavourable conditions of post-combustion flue gas at the stack.

Solubility and pH of CO2 in HJO

The solubility of CO2 in water is 0.9 vol CO2/vol H2O or 0.0007 mol CO2/mol H2O at 293.15 K.13,14 CO2 forms weak carbonic acid when dissolved. This dissolution, however, may reduce the pH of water to as low as 5.5.12 This finding is important, as it confirms that CO2 acts as an acid in acid-base reactions, which is vital information in the selection of solvents or sorbents which may be used to absorb or adsorb CO2.

Gas absorption using solvents

The acidic nature of dissolved CO2 in water dictates the types of physical and chemical solvents that would potentially be successful for efficient CO2 absorption. Applicable chemical solvents include amine solvents and solutions, which result in CO2 absorption by zwitterion formation and easy deprotonation by a weak base.11 Promising potential physical solvents include chilled ammonia, Amisol and Rectisol solvents,2 and ionic liquids which consist purely of cations and anions. Huang and Rüther15 discovered that a Lewis acid type interaction occurs between CO2 and anions, with CO2 acting as a Lewis acid and anions acting as a Lewis base.

The selective absorption of CO2 can be achieved by passing the flue gas through an absorber through which solvent flows counter-currently. CO2 is selectively absorbed into the solvent and exits through the bottom, while other flue gas components are passed out through the top of the absorber (Figure 3).



The solvent loaded with CO2 is then heated and sent to a stripping column where desorption occurs. CO2 is released, while the unloaded solvent is recycled to the absorber.

The advantage of this strategy is that the process is well developed as it is already in use for other gas treatment requirements such as desulphurisation and denitrification processes. There are many possible solvents and solvent mixtures that are under investigation for CO2 absorption, including amine and carbonate solvents, as well as ionic liquids.

The disadvantage is the high energy penalty associated with solvent regeneration in the stripping column. CO2 absorption increases with decreasing temperature, requiring flue gas to be cooled for CO2 absorption, as flue gas is available at a relatively high temperature of up to 413 K.16 However, thereafter, the loaded solvent needs to be heated in the stripping column to release CO2 and recycle the solvent. There is ongoing research on finding suitable solvents that are easily regenerated with a much lower energy penalty.

Pilot plants for processes of this type have already been set up in Austria and the Netherlands in 2008.17,18 South Africa's first CO2 capture plant that would likely include solvent absorption is scheduled to be set up by 2020, pending the success of CO2 injection projects19 (Surridge T, 2011, oral and written communication, November 28).

CO2 capture using dry regenerable sorbents

Figure 4 illustrates a sorbent adsorption process. Flue gas is first cooled and then sent to a carbonation reactor, which is a packed or fluidised sorbent bed reactor. CO2 is absorbed or adsorbed into the sorbents. This process may be physical or reactive. The sorbent, now loaded with CO2, is then transferred to a regenerator where it is heated to release the CO2. The sorbent is then recycled to the carbonation reactor.16



Packed bed reactors are popular for inherently porous activated sorbents while sorbents occurring as pellets, flakes, or fine particulate matter are used in a fluidised bed reactor. The process may operate in continuous or batch mode, depending on the efficiency of solids handling and the CO2 removal capacity of the process.

Common sorbents under investigation for CO2 capture include activated coal, sodium carbonate, potassium carbonate and calcium carbonate.16 CO2 capture is efficient even at low CO2 concentrations in the flue gas. Depending on the sorbent and process design, lower regeneration energy requirements can be achieved than those from absorption using amine solvents.14,17

The low attrition resistance of many sorbents is a fundamental setback to their implementation as a CO2 capture technique.2,18 While single-cycle results seem promising, many sorbents are not robust enough to be used in multi-cycle operation with conventional solids handling techniques. Sorbent pellets may erode or become caked and lose shape. High water content in the flue gas results in further attrition and sorbent caking. Moreover, the expensive nature of solids handling, including conveyor belts and compressed air blast loops which require maintenance, also reduces the feasibility of using sorbents as a CO2 capture technique.

Research, especially on the introduction of additives and sorbent supports and hybrid processes that combine sorbents with solvents, is being conducted to overcome the current challenges experienced with the use of sorbents.20 Details of a pilot plant set-up and usage are provided by Manovic et al.21 who utilised a fixed bed reactor. Fluidised bed pilot projects have also been considered in Canada and Korea.22,23

CO2 molecular size

The CO2 properties presented by the Asia Industrial Gases Association13 show that the CO2 molecule - a carbon atom double bonded to two oxygen atoms - is compact. Also, importantly, the molecule is linear in shape, with a bond length of 116.18 pm, making the molecule approximately 232 pm in length. Figure 5 provides an illustration of the molecule.



By comparison, the diatomic O2 molecule has a bond length of 120.8 pm, water has a bond length of 102.9 pm and N2 has a bond length of 109.7 pm.24 The size and linear shape of the CO2 molecule in relation to other flue gas components, as well as other properties such as dipole moment and polarisability, facilitates not only the use of sorbents, but also the use of conventional membrane filtration systems, enzymatic membranes and metal organic frameworks to filter out CO2 from smaller molecules of various components of flue gas.

Membrane separation

Figure 6 illustrates a typical membrane contactor. Flue gas enters into a membrane separation unit. CO2 selectively permeates through the membrane while other flue gas components do not. Flue gas passes out as stack gas, while CO2 is recovered and compressed on the other side of the membrane.



While membranes can be used on their own, increased efficiency is noted when solvents are used as a sweep fluid to accelerate mass transfer and recover CO2 on the other side of the membrane. Some solvents, such as ionic liquids, are combined into the membrane pores to increase CO2 permeability through the membrane.2

Common membrane material includes zeolite, ceramic, polymer and silica. More fragile membranes are supported by alumina to increase their robustness. Depending on the type of filtration unit, the process can operate in batch or continuous mode.

The advantage of membranes is that CO2 can potentially be recovered at high purity. Membrane units are well developed and there is high scope of study regarding membrane types and solvent combinations. If no solvent is used, then solvent regeneration and recycling is not required.

The challenge in implementing membrane separation for CO2 capture is the high pressure that the process demands. The flue gas needs to be compressed before undergoing filtration in order to achieve a high CO2 removal rate, which amounts to a high energy penalty. Moreover, many types of membrane material cannot satisfy optimum CO2 permeability and selectivity constraints and are not robust enough for long-term operation. Satisfying these requirements forms part of ongoing research.

A pilot plant in the Netherlands which accommodates CO2 capture using membranes combined with solvents was constructed in 2008.18

Enzyme-based systems

Instead of using conventional membranes as previously described, enzymes can be used as a liquid membrane suspended between hollow fibre supports for rigidity. As shown in Figure 7, flue gas passes through the liquid membrane. CO2 is hydrated and permeates as carbonic acid (HCO3-) at a faster rate than N2, O2 and other flue gas components. CO2 is recovered under pressure or using a sweep gas on the other side.2



A popular enzyme for CO2 capture is carbonic anhydrase. CO2 recovery with this technique can potentially be as high as 90%.2 About 600 000 molecules of CO2 are hydrated by one molecule of carbonic anhydrase.26 A further advantage is that the heat of absorption of CO2 into carbonic anhydrase is comparatively low.

Disadvantages include limitations at the membrane boundary layers, long-term uncertainty, and sulphur sensitivity of the enzyme,26 prompting ongoing research on new enzymes.

Research in this technique has not gone beyond laboratory studies on CO2 permeability and selectivity.26,27

Metal organic frameworks

Metal organic frameworks (MOFs) are hybrid organic/inorganic structures containing metal ions geometrically coordinated and bridged with organic bridging ligands28 (Figure 8). This arrangement increases the surface area for adsorption, enabling them to be used as sorbents or as nanoporous membranes.



MOFs possess much potential for CO2 capture because there are hundreds of possible MOFs that can be developed using various combinations of metal ions and organic ligands. They can be tai